NY Times letter-boxed solver

letter-boxed-solver.py

import json
import sys
import time

# Lexicon must be a JSON array of strings.
LEXICON_PATH = "lexicon.json"

if len(sys.argv) != 5:
    name = sys.argv[0]
    print(f"Usage  : {name} <side 1> <side 2> <side 3> <side 4>")
    print(f"Example: {name} ERP SBW OID UHC")
    sys.exit(-1)

sides = [side.lower() for side in sys.argv[1:]]
all_sides_string = "".join(sides)
full_letter_set = set(all_sides_string)

# Create a map from each letter to each playable letter, i.e. the letters on
# the other 3 sides.
side_map = {
    sides[0][0]: set(sides[1] + sides[2] + sides[3]),
    sides[0][1]: set(sides[1] + sides[2] + sides[3]),
    sides[0][2]: set(sides[1] + sides[2] + sides[3]),

    sides[1][0]: set(sides[0] + sides[2] + sides[3]),
    sides[1][1]: set(sides[0] + sides[2] + sides[3]),
    sides[1][2]: set(sides[0] + sides[2] + sides[3]),

    sides[2][0]: set(sides[0] + sides[1] + sides[3]),
    sides[2][1]: set(sides[0] + sides[1] + sides[3]),
    sides[2][2]: set(sides[0] + sides[1] + sides[3]),

    sides[3][0]: set(sides[0] + sides[1] + sides[2]),
    sides[3][1]: set(sides[0] + sides[1] + sides[2]),
    sides[3][2]: set(sides[0] + sides[1] + sides[2]),
}

def can_play(word):
    for i, letter in enumerate(word):
        if letter not in full_letter_set:
            return False
        if not i:
            continue
        if letter not in side_map[word[i - 1]]:
            return False
    return True

start_time = time.time()

lexicon = None
with open(LEXICON_PATH) as w:
    lexicon = json.load(w)

playable_words = [word for word in lexicon if can_play(word)]

# Create a mapping from a letter to all playable words starting with that
# letter.
playable_words_map = { letter: [] for letter in full_letter_set }
for word in playable_words:
    if word[0] in full_letter_set:
        playable_words_map[word[0]].append(word)

candidate_count = 0
for first_word in playable_words:
    last_letter = first_word[-1]
    second_word_candidates = playable_words_map[last_letter]
    for second_word in second_word_candidates:
        if set(first_word + second_word) == full_letter_set:
            candidate_count += 1
            print(f"Candidate: {first_word}, {second_word}")

elapsed = time.time() - start_time

print(f"\n{candidate_count} candidates found in {elapsed:.2f} secs.")